Brain damage due to intrauterine/perinatal virus infection is a significant congenital health problem. In addition, the CNS continues to develop during the first year of postnatal life, providing for continued susceptibility of the infant CNS to damage by viral infection after birth. Since the developing nervous system is uniquely sensitive to damage following virus infection, administering neurovirulent vaccines to infants also places the child's CNS at increased risk for injury. However, little is known about specific cellular components that support preferential replication of neurovirulent wild type and vaccine viruses in neuronal cells. Wild type mumps virus, and some strains of mumps vaccine virus (Urabe Am9, Leningrad 3), are among the most neurotropic of the early childhood viruses, and new MMR combinations continue to be proposed that include new strains of mumps vaccine virus. Thus, identification of the specific cellular characteristics that lead to enhanced susceptibility to damage by neurovirulent viruses will lead to the development of molecular biological, in vitro and/or small animal models of neurovirulence. Development of these non-simian models will lead to cost saving and improved predictability of neurovirulence testing. Identification of intracellular and cell surface molecules responsible for susceptibility to neurovirulent wild type and vaccine viruses will also provide information to help genetically engineer non-neurovirulent vaccine virus strains. Notably, information obtained in these studies about mumps virus vaccines will likely be useful in generalizing to other potentially neurovirulent vaccines (e.g., measles, Japanese encephalitis). Previous studies comparing virus growth in neuronal and non neuronal cell lines have suffered from problems related to the comparison of virus replication in cells of different lineage, strain and, sometimes, species. We have a human cell line (NT2) that can be differentiated from fibroblastic to neuronal cell type using retinoic acid. By comparing virus replication in these cells before and after differentiation to neuronal cell type, we can identify specific cell features important in neurovirulence using only one human cell line. Before and after differentiation, NT2 cells will be infected with parent or neurovirulent variants of non-neurovirulent (e.g., Jeryl Lynn vaccine), or parent or non-neurovirulent variants of neurovirulent (e.g., Kilham wild type) vaccine and wild type mumps virus. Differences in virus replication (e.g., binding of virus, titer of infectious virions, and viral proteins and RNA) will be compared between wild type and vaccine viruses and their variants before (fibroblastic) and after (neuronal) differentiation. Cell surface analysis, cell fractionation experiments and gel shift analysis will be used to identfy specific cellular components that interact with neurovirulent viruses before and after neuronal differentiation of cells. Once specific neuron-linked cell components that support neurovirulent virus replication are identified, we will identify the viral sequences responsible for this interaction. Using this information, virus vaccines can be developed with altered neurovirulence sequences to generate safer vaccines.